Certain embodiments of the present invention are directed to computer technology. More particularly, some embodiments of the invention provide systems and methods for signal processing. Merely by way of example, some embodiments of the invention have been applied to audio signals. But it would be recognized that the invention has a much broader range of applicability.
Two main factors, namely packet loss and latency, usually affect the quality of audio signals over an IP network. As a result of time-varying network conditions, the delay introduced by the processing at different nodes and/or terminals often leads to random time delay and packet loss. In order to improve network audio quality, packet loss compensation is often adopted to compensate for lost audio packets.
Currently, there are many audio packet loss compensation methods, such as: retransmission by a transmission terminal (i.e., to retransmit data packets upon confirmation that the data packets are lost), interleaving by a transmission terminal (i.e., to adjust a time sequence of data packets to be sent to reduce the impact of unexpected packet loss), silence substitution by a decoding terminal (i.e., to use silence to substitute lost data packets), data splicing (i.e., to directly splice together audio data before and after the lost data packets), packet duplication (i.e. to use an immediately preceding data packet to replace a current lost data packet), waveform interpolation (i.e., to use preceding and succeeding data packets to recover the lost data packets through interpolation), and model-based recovery (i.e., to construct a speech model or an audio model to recover the lost data packets).
The above-noted conventional methods have some disadvantages. For example, retransmission by the transmission terminal often needs retransmission of the entire lost data frame which usually greatly increases network transmission delay and end-to-end latency. Other above-noted methods often negatively affect the quality of reconstructed audio data.
Hence it is highly desirable to improve the techniques for audio coding.